Abstract This project seeks to develop a wireless, minimally invasive bi-directional deep brain stimulation technology based on remote heating of magnetic nanoparticles. Reliably modulating the activity of specific neuronal populations is essential to establishing causal links between neural firing patterns and observed behaviors. Electrical stimulation, as well as its recent non-invasive alternatives, ultrasound and electromagnetic induction, do not discriminate between cell types and have limited spatial resolution. Genetic approaches such as DREADDs and optogenetics enable neural excitation and inhibition with exquisite precision in specific cell populations. However, they require long-term indwelling hardware (limiting clinical translation) or lack temporal resolution. In this project, we propose to evaluate a nanoparticle-based technology that can access the deep brain regions, excite and inhibit neurons, and be fully wireless after initial injection. The Anikeeva (MIT) and Pralle (SUNY Buffalo) groups have recently shown that heat dissipation by magnetic nanoparticles (MNPs) in alternating magnetic fields (AMFs) can trigger heat-sensitive capsaicin receptor TRPV1 and heat-sensitive chloride channel anoctamine 1 (ANO1), respectively. These, in turn, can depolarize or silence neurons, and we have preliminary evidence for effects both in vitro and in vivo. Finally, the Anikeeva group has made advances in nanomaterials chemistry that enables multiplexing: independent heating of multiple MNP types (implying control of multiple neighboring neural populations) using AMF with distinct amplitudes and frequencies. Our objective is to combine these technologies into a magnetothermal toolbox and demonstrate its ability to shape animal behavior, by manipulating a well-characterized midbrain reward circuit. We will refine the ANO1 inhibitory technology and demonstrate control of place aversion in mice (Aim 1), then merge this technology with TRPV1-facilitated excitation in context of magnetic multiplexing to show bi-directional control of place aversion/preference (Aim 2). From this proof of concept, Aim 3 seeks to demonstrate that the toolkit can also control a more complex behavior (gambling/ probabilistic reward learning) in a larger species (rat). We will carry out this project through a tightly integrated combination of expertise in nanoscale engineering (Anikeeva, Pralle), targeted neural modulation (Anikeeva, Pralle), behavior manipulation through midbrain modulation (Widge) and clinical psychiatric deep brain stimulation (Widge).